s-waydriver
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates long-term coordination by persisting state, including external discovery results, into markdown files that guide subsequent agent actions.
- Ingestion points: Data enters the context via discovery results and repository evidence as described in SKILL.md and references/discovery.md.
- Boundary markers: The instructions do not provide explicit delimiters to isolate external content from the agent's control logic within the persisted records.
- Capability inventory: The skill possesses file-write capabilities for project documentation and the .scratch/ directory.
- Sanitization: No protocols are described for sanitizing or filtering external content before interpolation into state records.
- [NO_CODE]: The skill consists entirely of markdown documentation and configuration with no executable scripts or binaries.
Audit Metadata